pdata: Helper function for analyzing the performance of estimators

In arsenionhacolo/InferenceBEAGSD: Inference for Binary Endpoint Adaptive Group-Sequential Designs

Description

It takes the results produced by AnalyzeSimonDsgn or AnalyzeSimonDsgnAdaptN and produces a dataframe containing bias, mean square error and variance of the estimators.

Usage

pdata(t, design, stop, success, replicates)

Arguments

t	Dataframe containing results produced by AnalyzeSimonDsgn or AnalyzeSimonDsgnAdaptN.
stop	Taking value "yes", "no" or "both", indicating that only trials that stopped, continued or both were analyzed.
success	Taking value "yes", "no" or "both", indicating that only trials that were successful, unsuccessful or both were analyzed.
replicates	Number of trials analysed. It is equal to the number of rows in t.

Details

It is a helper function for AnalyzePerformanceSimon. It also calculates the power and the expected sample size (EN) where applicable.

Value

Dataframe containing bias, mean square error and variance of the estimators.

Author(s)

Arsenio Nhacolo

See Also

AnalyzeSimonDsgn, AnalyzeSimonDsgnAdaptN and AnalyzePerformanceSimon.

Examples

## Not run: 
rslt <- read.csv("ResultsOptimalDesign.csv")
nrep <- nrow(rslt)
t <- rslt
presult <- pdata(t, "Optimal", "both", "both", nrep)
t <- rslt[rslt$stop == 0,]
presult <- rbind(presult, pdata(t, "Optimal", "no", "both", nrep))

## End(Not run)

arsenionhacolo/InferenceBEAGSD documentation built on May 9, 2019, 4:10 a.m.